变精度粗糙集的约简及其参数β范围的确定

Attribute Reduction of VPRS and Determination of Its Parameter β Range

针对启发式算法获取最小约简的局限性,提出了一种基于变精度粗糙集理论的获得最小约简的新方法。该方法是将信息系统的所有属性约简以及每个约简的参数β范围结合起来考虑,从中选出所含属性个数最少而且参数β范围最大的约简,作为最小约简。同时给出了相应的求得所有约简及其参数β范围的算法,并且通过MATLAB进行了实现,结果表明以上算法对于条件属性较少的信息系统十分有效。最后通过算例说明所给方法的有效性及可行性。

A new method was proposed to find a minimal reduction, which was based on VPRS (variable precision rough set) theory and broken the limitation of heuristic algorithm. Both the attribute reductions in VPRS and the range of their parameter β were considered in the new method and the reduction with the minimal number of attributes and the maximal range of β were selected as a minimal reduction. Corresponding algorithms to get all reduction and parameter β were given and realized by MATLAB. The results show that these algorithms are more effective for the information system of less condition attributes. At last, an example was given to explain the validity and feasibility of algorithms.